IMPACT OF SEMANTIC TECHNOLOGIES TO HUMAN BEHAVIOR

MODELING

A Psychosocial Rationalization

Ana Bel

en Sanchez-Calz

on, Carlos Fern

andez-Llatas, Flavio Pileggi and Teresa Meneu

ITACA-TSB, Universidad Polit

ecnica de Valencia, Valencia, Spain

Keywords:

Semantic interoperability, Human behavior, Healthy habits, Psychology.

Abstract:

The promotion of healthy habits have lots of beneﬁts. In this way, the discovery of human health habits will

allow to experts, among others advantages, evaluate the accuracy of the promotion models. Nevertheless,

to design and develop a model that correctly predict and anticipate the behavior of an individual or a group

is a difﬁcult task, in the sense that the challenge of representing the behavior includes such diverse areas as

simulating the effects of making decision, modeling the cognitive processes that take place in making that

decision, or simulate the perception of motor skills. The design of an uniﬁed method to access the current

human behavior theories will facilitate the application of motivation technologies in an holistic way.

In this paper, a review about the factors involved on human behavior modeling, and the most important theories

on people health behavior is made and a ﬁrst attempt for the creation of a uniﬁed health behavior model is

presented.

1 INTRODUCTION

The discovery of the human behavior is one of the

hardest current challenges in multiple research ﬁelds.

The difﬁculties that scientist ﬁnd in their way are not

only the problem of detect what are the rules and fac-

tors that affects to human behavior, but also how that

knowledge can be represented.

The model of human behavior has been treated

from psychological and psychosocial research ﬁelds

from lots of years ago (Hochbaum, 1958), (Lewin,

1943). One of the ﬁelds where the human behavior

discovery has more applications is the healthy habits

promotion. The promotion of healthy habits have lots

of beneﬁts, not only in the health of individuals of cur-

rent society, but also in economical situation, cause

the large account of money that could be saved in

treatments for illnesses directly related with the un-

healthy habits of people. For that, the governments

year by year, apply an important part of their budget in

healthy habits promotion. Nevertheless, despite these

waste of money it is very difﬁcult to know what is the

exact impact of their promotion activities.

In literature, some emerging technologies in the

automatic human behavior modeling based on in the

pattern recognition research ﬁeld could be used to test

the capabilities of health promotion techniques ap-

plied by governments and health professionals. Com-

plex Event Processing (Wasserkrug et al., 2008) tech-

niques allows process events and discover complex

patterns among multiple streams of event data; Plan

recognition Models (Phua et al., 2009) allows the

alignment of habits of individuals to detect unhealthy

situations and evaluate the compliance to promotion

campaigns. Process Mining (A. K. A., Workﬂow

Mining) (Fern

andez-Llatas et al., 2009) techniques al-

lows detecting changes in the behavior of individuals

or people as a whole detecting if the promotion cam-

paigns has enough impact over their.

Nevertheless the huge amount of factors and dif-

ferent theories make difﬁcult to these techniques to

focalize what are the factors to be taken in to ac-

count. Although the different psychological and psy-

chosocial research lines seem to be theorically com-

plementaries it is needed an effort to create a com-

mon framework to condense the psychological theo-

ries. The use of ontologies to describe de concepts

and to provide an mechanism to semantically interop-

erate among them.

In this paper, a review about the factors involved

on human behavior modeling, and the most important

theories on people health behavior is made and a ﬁrst

attempt for the creation of a uniﬁed health behavior

model is presented.

547

Belén Sanchez-Calzon A., Fernández-Llatas C., Pileggi F. and Meneu T..

IMPACT OF SEMANTIC TECHNOLOGIES TO HUMAN BEHAVIOR MODELING - A Psychosocial Rationalization.

DOI: 10.5220/0003886705470552

In Proceedings of the 4th International Conference on Agents and Artiﬁcial Intelligence (IWSI-2012), pages 547-552

ISBN: 978-989-8425-95-9

 2012 SCITEPRESS (Science and Technology Publications, Lda.)

2 PSYCHOSOCIAL THEORIES IN

HUMAN BEHAVIOR

MODELING

Based on the human behavior factors previously de-

scribed the scientiﬁc community has implemented

some theories in order to explain in what way these

factors affects to human behavior in their Health mo-

tivation. The most important theories are the Health

Belief Model, the Theory of Reasoned Action, the So-

cial Action Theory and Self-Efﬁcacy Theory

2.1 Health Belief Model

This theoretical development was ﬁrst formulated by

Hochbaum (Hochbaum, 1958), and subsequently ex-

tended by Rosentock (Rosenstock, 1974) applied to

the explanation and prediction of a wide range of

health behaviors. The proposal is the explanation

of behavior by focusing on psychosocial variables,

from a body of psychological theories such as Field

Theory by Lewin (Lewin, 1943) (Lewin, 1946), Ex-

pected Value Theory, and the Theories of Decision

making Under Uncertainty. The initial hypothesis of

this model is in the following way: an individual does

not perform a healthy behavior if she dont have min-

imum levels of relevant motivation and information

to health, she sees herself as potentially vulnerable,

and she sees the disease (or risk) as threatening, she is

convinced of the effectiveness of the intervention, and

ﬁnd little difﬁculty in the practical action of healthy

behavior (Rosenstock, 1974)).

In the ﬁrst place, according to the Health Belief

Model, the probability of making a healthy behavior

depends on the subjective state of intention of the in-

dividual to do so. The intention is determined by the

threat of the disease in question (according to the be-

lief of the individual). The perceived threat is deter-

mined by: the perceived susceptibility for the disease,

perceived severity of the consequences of suffering

this disease, and keys to trigger action to an appropri-

ate health behavior. These keys may come from inter-

nal sources (symptoms etc.) or external (interactions

with other, media etc.) (Becker and Maiman, 1982).

Secondly, the probability that an individual devel-

ops and maintains a healthy behavior depends on an

assessment that makes such behavior in terms of prac-

ticability and effectiveness, countered with the per-

ceptions of the physical, economic costs and other

barriers involved in the proposed action. These per-

ceived costs have been assessed in different ways in

different studies: in terms of security on the effective-

ness of the treatment prescribed, the patient satisfac-

tion about the communication with the doctor, patient

satisfaction about matters such as the way in which

the health organization provides medical care, etc.

Third, the threat perception triggered by some key

event starting the process of making healthy behav-

ior. Health Belief Model is based on the premise that

socio-demographic, structural and individual factors

may inﬂuence health behaviors. However, it consid-

ered that these variables act through their effects on

individual health beliefs, and not so direct causes of

wholesome actions (Becker and Maiman, 1982).

Most explanatory research on the effectiveness of

Health Belief Model are retrospective and it isnt pos-

sible to determine, from them, if the health belief

existed before or it was created after the conclusion

of the studied behavior. Its an appropriate model as

an explanatory model of preventive behavior, and an

inappropriate model to explain behaviors oriented to

health promotion.

2.2 Theory of Reasoned Action

Fishbein (M. and Ajzen, 1975) developed the The-

ory of Reasoned Action, and then further developed in

the Theory of Planned Action (Ajzen, 1987) (Ajzen,

1988), which attempts to explain the behaviors that

are under conscious control of individuals from differ-

ent determinants that precede and explain them. For

the authors, the immediate determinant of behavior is

not the attitude itself, but the intention to do. In turn,

behavioral intention has two forerunners that explain

it: a single precursor, that is, the attitude about the be-

havior, and another social and collective that refers to

the socio-cultural context of the individual, named as

the subjective norm (M. and Ajzen, 1975). Both atti-

tude and subjective norm are determined by other fac-

tors that precede them, helping to better understand

the behavior. As regards the attitude, it is determined

by each of the beliefs that person has to an object and

the assessment based on those beliefs. This assess-

ment represents the affective component of attitude,

determining the motivation and the strength of behav-

ioral intention.

According to this theoretical construct, beliefs

vary according to their origin, and can be formed from

different processes:

• Direct experience regarding the attitude object,

through which information is collected on the

characteristics of the object (a person, thing etc.).

The attitudes formed from this process have

greater strength and they are more resistant to

change.

• Indirect experience regarding the attitude object,

which brings the same features of that object by

the similitude to other objects with which it has

ICAART 2012 - International Conference on Agents and Artificial Intelligence

548

had previous direct experience. Such beliefs are

called inferential.

• The information collected from other (media,

family, friends, etc.). Information is accepted as

real, unless it contradicts the beliefs formed from

the direct or indirect experience (M. and Ajzen,

1975).

Regarding to the subjective norm, it has deter-

mined by the perception of the beliefs that others have

about the conduct that the individual must perform

and, moreover, it is also determined by the individ-

ual’s motivation to meet the expectations that others

have about her. Its a differential process of beliefs

formation that contributes to each of the beliefs have a

particular weight and value according to each individ-

ual and the attitude object. Outstanding attitudes, to-

gether with their evaluation, will better predict the be-

havior intention (M. and Ajzen, 1975). Furthermore,

knowledge about the speciﬁc beliefs of what others

persons think of each of the speciﬁc behaviors (for

example smoking, physical exercise, healthy eating,

etc.) will Inﬂuence the intention to carry out or not

a general behavior (maintaining a healthy lifestyle),

depending on the motivation to please.

However, not all behaviors are consciously con-

trolled by the individual, because there are many situ-

ations in which contingency may occur, or in which it

requires skills or resources that could interfere with

the intention of performing a behavior. Its neces-

sary a third determinant of behavioral intention, the

perceived control (Ajzen, 1987) (Ajzen, 1988)). Al-

though the individual has a favorable attitude toward

a behavior, the probability of carrying it out will de-

pend on, among other factors, the perception of con-

trol by the individual about her behavior. A person

may have a favorable attitude toward health care, for

instance to give up smoking, but if that individual per-

ceives that she has low ability to control and to quit

smoking, either because she considers she hasnt ca-

pacity enough, or because she believes that behav-

ior of other people can interfere with her decision to

quit, this healthy behavior will not be performed. The

perception of control is a factor consisting of inter-

nal variables (perceived ability, skill of action etc.),

and external variables (opportunity to action, obsta-

cles, time, cooperation, etc.). This is a determinant

that helps to improve forecasting and modeling of the

behavior (Ajzen, 1987).

The Theory of Reasoned Action doesnt take into

account factors such as attitudes towards goals, per-

sonality characteristics, sociodemographic variables,

social role, etc. For this model, these factors are exter-

nal variables that can inﬂuence behavior, but no nec-

essary relationship between them.

2.3 The Social Action Theory

This model proposed by Ewart (Ewart, 1991) presents

the individual as a self-regulatory system that actively

trying to achieve goals, and also as a feedback sys-

tem consisting of a set of serially arranged elements,

or stages: a ﬁrst input stage, which represents stimu-

lus ﬁeld and sets the targets. A second output stage,

or production of response, which provides the plans,

the selection and development of responses. And a

third stage of supervision that involves the consider-

ation of the consequences of action in relation to the

initial set of objectives. This model stresses the role

of social context in the development and maintenance

of healthy routines and habits; it provides the causal

structure linking the self-change processes in inter-

personal environments; and it also speciﬁes the social

and environmental inﬂuences that make possible per-

sonal change.

According to Social Action Theory, preventive in-

terventions involve the creation of protective habits in

the form of routines behavioral sequences that reduce

personal risk. The actions are guided by their conse-

quences in a control loop, and variations in the results

produce compensatory behavioral adjustments. The

result is a steady state-action, but constantly ﬂuctu-

ating. The starting point for intervention is an anal-

ysis of the relations between the harmful or beneﬁ-

cial behaviors to health and its effects experienced.

This is an analysis that allows extracting those as-

pects in which behaviors are more accessible to pre-

vention, and it also allows extracting effective pro-

cedures for the design of new schemes that promote

health (Ewart, 1991).

The Social Action Theory remarks that per-

sonal action schemes are socially interconnected to

schemes of close social environment (friends, family,

peers etc.) so they raise signiﬁcant obstacles to long-

term changes. Its necessary to extend the concept of

state-action focused individually to include social in-

terdependence as a determinant of a change in be-

havior. Close social relations mean that the patterns

of action of each person are interconnected, thereby

increasing the likelihood of someone trying to mod-

ify a routine, inﬂuencing and conditioning routines of

other individuals.

This model implies the existence of mechanisms

that enable people to make transitions from old states-

action to other new, causing a change. According to

this theory, attitudes and reinforcements do not deter-

mine and cause behavior. People motivate themselves

by taking into account the possible outcomes, assess-

ing their capabilities and creating goals that guide

and energize the solution of various problems. The

IMPACT OF SEMANTIC TECHNOLOGIES TO HUMAN BEHAVIOR MODELING - A Psychosocial Rationalization

549

health related behaviors are included, according to

the Theory of Social Action, in a set of norms of

behavior focused toward some important goal. This

set of schemas is composed of individual’s personal

projects, it refers to basic tasks (such as getting social

inﬂuence, acquire material resources, be accepted by

others, etc.), and it affects to protective rules of behav-

ior making people generate aimed objectives to eval-

uate their own conduct.

The schemas represent organized sets of behav-

iors that focus our attention to speciﬁc aspects of sit-

uations and scenarios. They drive the encoding of ex-

perience in memory over the long term. And ﬁnally

they provide routines to make tasks. These ways of

knowledge integrate capabilities which allow to con-

sider alternative goals, and to create new action strate-

gies. Capacities are the mechanisms through which

physical and social contexts affect to self-regulatory

conducts. Cognitive control schemes inﬂuence the

choices of behavior, increasing trust in the individ-

ual’s ability to maintain or change a behavior. The

context or environment in which people live adjust

the personal capacities and the social relations, af-

fecting the establishment of goals, the consideration

of opportunities for action, and the design of rele-

vant health strategies. The physical characteristics

of the environment determine the access to material

resources, and they also inﬂuence in the behavioral

strategies. Social relationships involve a set of bene-

ﬁts, expectations and obligations that likewise inﬂu-

ence on the objectives aimed. In addition, social re-

lationships provide behavior models that facilitate or

inhibit the action guidelines (Rodriguez, 2001).

2.4 Self-efﬁcacy Theory

Bandura (Bandura, 1977) (Bandura, 1988) developed

the Self-efﬁcacy Theory to explain human behavior

and the factors involved in motivation. Self-efﬁcacy

can be deﬁned as the evaluation of the own personal

abilities in view of the possibility of action. There are

different processes that contribute to the conﬁguration

and development of self-efﬁcacy. They are elements

that describe how certain characteristics, both inter-

nal and external to the individual, which also help to

shape the beliefs that she has about what is able or

unable to do:

• Direct experience is the main source of formation

regarding the self concept that an individual has.

In this sense, the consequences obtained after ex-

ecuting the behavior make it possible to report

about one’s ability to carry out the same behav-

ior, and if is probable to control the situational

variables in which this behavior has to be done.

Thus, the experience and consequences obtained

contribute to the formation of self-concept, and

contribute also to develop of personal self-worth

feeling, both required to deal in different situa-

tions.

• Learning by observation is also a source of valu-

able information. To observe and analyze the con-

sequences that any action cause in another indi-

vidual executing such action can lead both to in-

hibit and to promote the own action, depending

on positive or negative assessment of the observed

consequences in other individuals. Its a process of

social comparison in which the individual shape

her perception of her own capabilities to deal with

different situations.

• Persuasive Messages. The attempts of people in

the environment to convince and encourage the

individual to take a speciﬁc action can provide se-

curity and support, in order to really produce a be-

havior. However, persuasion is a weaker resource

than previous ones, and it will vary depending on

different variables such as, for example, certain

features related to the personality of the individ-

ual who tries to persuade, the credibility that indi-

vidual has to pass on, or her ability to get another

individual execute the action.

• Physiological activation may also be an impor-

tant modulator of the capabilities that a person be-

lieves she has, can inﬂuencing in the process of

self valuation. According to (Bandura, 1988), the

information provided by the psycho-physiological

activation inﬂuences the perceived effectiveness

through assessment processes. Therefore, when

such activation happens in the view of the possi-

bility of executing a behavior, the individual eval-

uates different factors, stressing the sources that

cause the activation of behavior, the intensity of

activation, the circumstances in which that activa-

tion happens, and the way in which the activation

inﬂuences over the efﬁciency. When physiologi-

cal activation is successful after the execution of

a behavior on previous experience, this activation

is considered by individuals as a facilitator of the

action, while, in opposition, when it is unsuccess-

ful, it is considered as harmful or inhibitor of the

action (Bandura, 1988).

People increasingly develop their self-efﬁcacy

perception based on executed behaviors, based on

said explanation in their environment, and on the rein-

forcements that other people around them do respect

the conduct has been made.

ICAART 2012 - International Conference on Agents and Artificial Intelligence

550

Figure 1: General Architecture of the uniﬁed semantical model.

3 TO AN UNIFIED SEMANTICAL

MODEL TO REPRESENT THE

PSYCHOLOGY OF HEALTH

BEHAVIOR

In literature, there are some examples of how the psy-

chology of patients is taken into account to help the

motivation of healthy behaviors (S. and Price, 2009),

(del Hoyo-Barbolla E. et al., 2006), (Hu and Sundar,

2010), (Gustafson et al., 2011). The research in that

ﬁeld has been based on psychological and psychoso-

cial research ﬁelds that are limited to the theoreti-

cal framework that the authors follows. This is be-

cause, in practice, the health social psychology scien-

tiﬁc community follows those models in parallel, rep-

resenting the same knowledge from different points

of view. Nevertheles, although described models have

differences in the theoretic perspective, they seems to

be complementary. In ﬁgure 1 a general approach to a

collaborative model of the different research theories

is presented.

In that schema, the different models are intercon-

nected according to their main hypothesis in order to

resolve the psychological causes of health behavior

The Health Belief model, mainly based on per-

sonal motivation can be enriched by results achieved

by Theory of Reasoned Action. In this case, the per-

sonal motivation available data might be improved by

the attitude and Subjective Norm of the second theory.

At the same time, according the Self-Efﬁcacy Theory

the attitudes can be continuously evolving depend-

ing on the health Behavior results perceived by the

user. Finally, the Subjective Norm user proﬁle can be

enriched with the thesis of Social Motivation, Social

Context and Relatives Belief in Social Action Theory

The general concepts that these models handle are

equivalents. This allows that the result of the research

on one of the lines can be useful as an entry point

for the rest. In this way, the formal representation of

the results using ontologies allows their use not only

in the same research ﬁeld but also in the others. In

addition, based on that uniﬁed model and representing

the knowledge of the thesis of those different theories

in a semantically tagged concepts it was possible to

perform automatic reasoning over the whole theories

in the psychology of health behavior research ﬁeld.

The creation of an uniﬁed semantic framework to

represent the thesis of Psychology of health behav-

ior will suppose a great advantage in order to provide

a uniﬁed view of the knowledge in that ﬁeld where,

psychologist, sociologist and health professionals not

only will allow to share their knowledge, but also will

be able to enrich their thesis with the direct result of

the investigation of their colleagues.

4 CONCLUSIONS

The use of ontologies to allow the semantic interop-

erability among systems that represent the different

health behavior models will enable de creation of an

uniﬁed model that explain the whole health behav-

ior of people. Experts will be able to share not only

in theory but also their knowledge might be used to

reason over the data available in order to create basic

IMPACT OF SEMANTIC TECHNOLOGIES TO HUMAN BEHAVIOR MODELING - A Psychosocial Rationalization

551

knowledge that could be used by other research ﬁelds.

These systems can offer a better knowledge that

can be used in conjunction with emerging pattern

recognition theories in order to validate the use of

health promotion techniques.

REFERENCES

Ajzen, I. (1987). Attitudes, traits, and actions: Disposi-

tional prediction of behavior in personality and social

psychology. Advances in Experimental Social Psy-

chology, 20(C):1–63. cited By (since 1996) 183.

Ajzen, I. (1988). Attitudes, Personality, and Behavior. The

Dorsey Press. cited By (since 1996) 1506.

Bandura, A. (1977). Self-efﬁcacy: Toward a unifying

theory of behavioral change. Psychological Review,

84(2):191–215. cited By (since 1996) 6537.

Bandura, A. (1988). Self-regulation of motivation and ac-

tion through goal systems In Cognitive perspectives

on emotion and motivation. Kluwer Academic Pub-

lishers.

Becker, M. and Maiman, l. (1982). Models of Health-

Related behavior in Handbook of Health, Health Care

and the Health Profession. Free Press.

del Hoyo-Barbolla E., R., K., Arredondo, M., and Ortega,

M. (2006). A new perspective in the promotion of e-

health. Studies in health technology and informatics,

124:404–412. cited By (since 1996) 3.

Ewart, C. (1991). Social action theory for a public health

psychology. American Psychologist, 46(9):931–946.

cited By (since 1996) 124.

Fern

andez-Llatas, C., L

azaro, J. P., and Bened

ı, J. M.

(2009). Workﬂow mining application to ambient

intelligence behavior modeling. In Universal Ac-

cess in Human-Computer Interaction, volume 5615 of

Lecture Notes in Computer Science, pages 160–167.

Springer.

Gustafson, D. H., .and Bret R. Shaw, M. G. B., Isham,

A., McTavish, F., Richards, S., Schubert, C., Levy,

M., and Johnson, K. (2011). An e-health solution for

people with alcohol problems. Alcohol Research &

Health, 33:327–337.

Hochbaum, G. (1958). Public participation in medi-

cal screening programs: A sociopsychological study.

Technical report, PHS publication no. 572. Washing-

ton D.C.: U.S. Government Printing Ofﬁce.

Hu, Y. and Sundar, S. S. (2010). Effects of online health

sources on credibility and behavioral intentions. Com-

munication Research, 37(1):105–132. Cited By (since

1996): 9.

Lewin, K. (1943). Deﬁning the ’ﬁeld at a given time.’. Psy-

chological Review, 50(3):292–310. cited By (since

1996) 116.

Lewin, K. (1946). Action research and minority problems.

Journal Social Issues, 2(4):34–46.

M., F. and Ajzen, I. (1975). Belief, Attitude, Intention and

Behavior: An Introduction to Theory and Research.

Addison-Wesley. cited By (since 1996) 6992.

Phua, C., Foo, V. S.-F., Biswas, J., Tolstikov, A., Aung,

A.-P.-W., Maniyeri, J., Huang, W., That, M.-H., Xu,

D., and Chu, A. K.-W. (2009). 2-layer erroneous-plan

recognition for dementia patients in smart homes. In

Proceedings of the 11th international conference on e-

Health networking, applications and services, Health-

com’09, pages 21–28, Piscataway, NJ, USA. IEEE

Press.

Rodriguez, J. (2001). Psicologia Social de la Salud. Sinte-

sis Psicologia.

Rosenstock, I. (1974). Historical origins of the health belief

model. Health Education Monographs, 2.

S., C. and Price, A. (2009). Healthy living with persua-

sive technologies: Framework, issues, and challenges.

Journal of the American Medical Informatics Associ-

ation, 16(2):171–178. cited By (since 1996) 10.

Wasserkrug, S., Gal, A., Etzion, O., and Turchin, Y. (2008).

Complex event processing over uncertain data. In

Proceedings of the second international conference

on Distributed event-based systems, DEBS ’08, pages

253–264, New York, NY, USA. ACM.

ICAART 2012 - International Conference on Agents and Artificial Intelligence

552